The present invention relates to a fuzzy-neuro method for discriminating optical disk, especially to a fuzzy-neuro method using a simple architecture to discriminate optical disk type for an optical disk driver.
The conventional optical disks have many different kinds of formats such as CD, CD-R, CD-RW, single-layer DVD (DVD5), double-layer DVD (DVD9), DVD-R and DVD-RAM. The different kinds of optical disks have different data formats and physical characteristics. Therefore, the optical disk driver read different kinds of optical disks in different schemes. For example, the optical disk driver uses 780 nm laser for reading CD and 650 nm laser for reading DVD, some of the signals read from optical disk even requires amplification. The optical disk driver compatible to a plurality of kinds of optical disks needs an optical disk discriminating step to determine suitable accessing scheme.
The conventional optical disk driver uses the electrical signals such as focus error (FE), sub-beam add (SBAD) or RF envelope (RFENV) to discriminate the disk type. The electrical signals are converted from the optical signals detected by the pick-up head thereof. The optical disk driver sets different threshold levels corresponding to those electrical signals and uses bi-sectional method to compare each electrical signal with corresponding threshold level. The optical disk driver can discriminate the disk type by the comparison results of those electrical signals. However, bi-sectional method is not immune to measurement error and the discrimination may be deteriorated by the measurement error.
It is the object of the present invention to provide a fuzzy-neuro method using a simple architecture to discriminate optical disk type for an optical disk driver.
To achieve above object, the present invention provides a fuzzy-neuro method to discriminate optical disk type by using a neuro network. The method fuzzy quantizes a plurality of electrical signals measured from optical disks of various formats and establishing a plurality of characteristic signal sets (Wcd, Wcdr, Wcdrw, Wdvd5 and Wdvd9) corresponding to the optical disks of various formats (including CD, CD-R, CD-RW, DVD5, DVD9). The electrical signals include focus error (FE) signal, sub-beam add (SBAD) signal, and RF envelope (RFENV) signal. The plurality of electrical signals from an optical disk to be discriminated is measured to obtain an input signal set. The signal set is sent to an input layer of the neuro network and the characteristic signal sets are stored in a distance layer of the neuro network. The distance layer of the neuro network computes the distances (dcd, dcdr, dcdrw, ddvd5 and ddvd9) of the input signal set with respect to the characteristic signal sets stored in the distance layer of the neuro network. The distances of the input signal set with respect to the characteristic signal sets are sent to a distance value compare block of the neuro network. The distance value compare block select the minimal one among the computed distances; the characteristic signal set with minimal distance determines the type of the optical disk under question.
The various objects and advantages of the present invention will be more readily understood from the following detailed description when read in conjunction with the appended drawing, in which: